Three years of support are requested to continue and complete data collection and analysis of a set of randomized universal intervention trials designed to prevent Conduct Disorders (CD) in first- and fifth- grade boys and girls. The strategy integrated a set of interventions across peer, classroom, and home settings, and was based on, a well-documented developmental model of CD. Initial and very preliminary results suggest that intervention did, in fact, have a significant immediate impact. The three years will also be used to collect follow-up data testing for true preventive impact, not only of CD--a condition resistant to treatment once it becomes chronic-- but of its serious correlates (e.g., substance abuse,, early sexual activity and parenthood, and accidental and self-inflicted injury) - The evolving multimeasure, multisetting data set will be used to test the theoretical models on which the trials were based, and to develop models for the development of CD in girls.